Hochschulschriftenserver der TH Wildau
Not a member yet
1283 research outputs found
Sort by
3D printed rotor blades for a research wind turbine: Aerodynamic and structural design and testing
No full textThis study combines the design, the 3D printing and the testing of a small 3-bladed wind turbine rotor for research and teaching purposes. The objective is the additive manufacturing of a rotor with a radius of one meter, as an alternative to subtractive methods, such as computerized milling. The blade design is developed using freely available software packages. The aerodynamic considerations include the airfoil selection, the calculation of the blade geometry and the simulation of the ultimate load cases. The structural considerations are focussed on the printable materials, the infill structures and the retrofit of a load-carrying spar. The rotor blades are 3D printed with the BigRep One at the maker space of the TH Wildau. The structural integrity of the prototype blade is tested in terms of the ultimate root bending moments and the centrifugal forces at the HTW Berlin. The aerodynamic run-up tests are performed at the large wind tunnel of the TU Berlin measuring the power curves. The successful prototype paves the way for follow-up projects, such as open field tests and the 3D printing of larger rotor blades
Machine Learning Methods in Predicting Patients with Suspected Myocardial Infarction Based on Short-Time HRV Data
No full textDiagnosis of cardiovascular diseases is an urgent task because they are the main cause of death for 32% of the world’s population. Particularly relevant are automated diagnostics using machine learning methods in the digitalization of healthcare and introduction of personalized medicine in healthcare institutions, including at the individual level when designing smart houses. Therefore, this study aims to analyze short 10-s electrocardiogram measurements taken from 12 leads. In addition, the task is to classify patients with suspected myocardial infarction using machine learning methods. We have developed four models based on the k-nearest neighbor classifier, radial basis function, decision tree, and random forest to do this. An analysis of time parameters showed that the most significant parameters for diagnosing myocardial infraction are SDNN, BPM, and IBI. An experimental investigation was conducted on the data of the open PTB-XL dataset for patients with suspected myocardial infarction. The results showed that, according to the parameters of the short ECG, it is possible to classify patients with a suspected myocardial infraction as sick and healthy with high accuracy. The optimized Random Forest model showed the best performance with an accuracy of 99.63%, and a root mean absolute error is less than 0.004. The proposed novel approach can be used for patients who do not have other indicators of heart attacks
VIGA - Virtual Instructor for General Aviation
No full textThe most common cause of incidents and accidents in aviation is linked to the category “Loss of Control Inflight” [1]. Remarkably in consequence this means that aircraft without any technical defect or such with manageable defects according to certification requirements are involved. The research project “Virtual Instructor for General Aviation” (VIGA) was aimed to validate an idea that addresses this problem by an entirely different approach. The idea can be best described by looking at the way a flight instructor takes decisions to intervene. A human pilot has an expectation of the maneuvers and the corresponding trajectories that can be flown in the future based on the present flight conditions. Decision making is based on the analysis of the consequences of the expectations. This approach is one of the key principles of the project, and is completely different to any known AFCS system.
Yet technical implementation requires considerable effort. Essentially it comprises a faster than real time simulation with an adequately accurate aerodynamic model of the particular aircraft in combination with a module to evaluate the results of the simulated exit trajectories. In consequence this idea requires an autopilot module capable of tracking the calculated three-dimensional trajectories which then resulted in the need to design and develop a completely new type of autopilot algorithm.
The objective of this project was to test and demonstrate principle functionality thereby also finding pathways determining future developments as well as to analyze system behavior. Therefore, the project was deliberately designed to help the pilot by depicting the solutions on the PFD. Direct intervention of the system with aircraft flight controls did not take place. This also raised the question of how to design an effective visual human interface.
Fortunately, all results proved to be very satisfying. The underlying idea could be validated and was demonstrated both in a simulation environment and in flight test. The path tracking algorithm was developed in a parallel project and also showed very satisfactory results, meeting all requirements
Extracellular Glycolytic Activities in Root Endophytic Serendipitaceae and Their Regulation by Plant Sugars
No full textEndophytic fungi that colonize the plant root live in an environment with relative high concentrations of different sugars. Analyses of genome sequences indicate that such endophytes can secrete carbohydrate-related enzymes to compete for these sugars with the surrounding plant cells. We hypothesized that typical plant sugars can be used as carbon source by root endophytes and that these sugars also serve as signals to induce the expression and secretion of glycolytic enzymes. The plant-growth-promoting endophytes Serendipita indica and Serendipita herbamans were selected to first determine which sugars promote their growth and biomass formation. Secondly, particular sugars were added to liquid cultures of the fungi to induce intracellular and extracellular enzymatic activities which were measured in mycelia and culture supernatants. The results showed that both fungi cannot feed on melibiose and lactose, but instead use glucose, fructose, sucrose, mannose, arabinose, galactose and xylose as carbohydrate sources. These sugars regulated the cytoplasmic activity of glycolytic enzymes and also their secretion. The levels of induction or repression depended on the type of sugars added to the cultures and differed between the two fungi. Since no conventional signal peptide could be detected in most of the genome sequences encoding the glycolytic enzymes, a non-conventional protein secretory pathway is assumed. The results of the study suggest that root endophytic fungi translocate glycolytic activities into the root, and this process is regulated by the availability of particular plant sugars
Entwicklung und Einsatz von Codegenerierungssoftware zur Effizienzsteigerung bei der Prozessanlagenautomatisierung
No full textDie Programmierung von SPS-Systemen erfolgt hauptsächlich händisch, obwohl große Teile des Programmcodes sehr ähnlich sind und sich häufig nur durch die eingesetzten Variablen unterscheiden. Der Beitrag befasst sich mit der Entwicklung und Testung einer Codegenerierungssoftware für das TIA Portal, um den Zeitaufwand der Programmierung zu reduzieren und Flüchtigkeitsfehler zu vermeiden
Metadaten von Verlagen in Repositorien : eine Analyse und Auswertung der Datenlieferungen des Open-Access-Lieferdienstes DeepGreen
No full textOpen-Access-Publikationen gewinnen in Bibliotheken eine immer größere Bedeutung. Bibliotheken haben aufgrund verschiedener Lizenzierungsmodelle die Möglichkeit diese Publikationen auf ihren Repositorien zweit zu veröffentlichen. Um diesen Veröffentlichungsvorgang zu vereinfachen, bibliothekarische Bestände zu erweitern und die Arbeit der Bibliotheken zu automatisieren, gibt es den Open-Access-Lieferdienst DeepGreen. Er beliefert über 60 verschiedene Repositorien mit den Publikationen von mehreren Verlagen.
Zur Weiterentwicklung des Lieferdienstes soll die Metadatenqualität verbessert werden, um den Aufwand für die beteiligten Repositorien weiter zu verringern. Die vorliegende Arbeit soll einen Einstieg für die Verbesserung der Metadatenqualität bieten. Neben einer Umfrage unter den DeepGreen Beteiligten werden auch die bisher vorliegenden Metadaten genauer beleuchtet. Hieraus werden Handlungsempfehlungen für die zukünftige Entwicklung der Metadatenqualität im Rahmen von DeepGreen erarbeitet
Composite materials for innovative urban farming of alternative food sources (macroalgae and crickets)
No full textFacing an inexorable growth of the human population along with substantial environmental changes, the assurance of food security is a major challenge of the present century. To ensure responsible food consumption and production (SDG 12), new approaches in the food system are required. Thus, environmentally controlled, sustainable production of alternative food sources are of key interest for both urban agriculture and food research. To face the current challenge of integrating food production systems within existing structures, multidisciplinary discourses are required. Here, we bring together novel technologies and indoor farming techniques with the aim of supporting the development of sustainable food production systems. For this purpose, we investigated the feasibility of 10 composite materials for their innovative use as structural support in macroalgal cultivation (settlement substrates) and cricket rearing (housing). Considering material resistance, rigidity, and direct material-organism interactions, the bio-based composite polylactic acid (PLA) was identified as a suitable material for joint farming. For macroalgae cultivation, PLA sustained the corrosive cultivation conditions and provided a suitable substrate without affecting the macroalgal physiology or nutritional composition (carotenoids and chlorophylls). For cricket rearing, PLA provided a suitable and recyclable shelter, which was quickly accepted by the animals without any observed harm. In contrast, other common composite components like phenolic resin or aramid were found to be unsuitable due to being harmful for the cultivated organisms or instable toward the applied sterilization procedure. This multidisciplinary study not only provides profound insights in the developing field of urban indoor food production from a new perspective, but also bridges material science and farming approaches to develop new sustainable and resilient food production systems
A comparison of SUMO’s count based and countless demand generation tools
No full textThere are already several tools available to generate traffic demand for the microscopic simulation suite SUMO. This paper focuses on setting up a simulation scenario for the peak hour in a small conurbation when there are vehicle counts available for the major streets. We describe tools which are part of SUMO or available as open source and compare their results with the real traffic counts as well as with the outcome of countless demand generation
Conception and implementation of an artificial neural network for situation recognition with humanoid robots based on several input channels
No full textIn dieser Masterarbeit soll eine Architektur für ein Künstliches Neuronales Netz erarbeitet werden, mit der mehrere Eingangskanäle eines humanoiden Roboters für eine Situationserkennung zusammengefasst werden können. Dazu werden zunächst die Grundbausteine für eine Situationserkennung am Beispiel des menschlichen Gehirns untersucht und die Erkenntnisse auf die Arbeitsweise Künstlicher Neuronales Netze übertragen.
Anschließend wird ein Konzept für eine Situationserkennung auf humanoiden Robotern erstellt. Dazu werden mögliche Situationen konstruiert und die Aufgabe der Situationserkennung so formuliert, dass sie mit einem Künstlichen Neuronalen Netz bearbeitet werden kann. Daraufhin werden mögliche Architekturen für dieses Netz verglichen und die am besten geeignetste ausgewählt.
Das erstellte Konzept soll im Rahmen dieser Arbeit auf dem Tisch-Roboter ROS-E umgesetzt werden. Dazu soll eine Trainings- und Testumgebung geplant und umgesetzt werden. Nachdem das entwickelte Künstliche Neuronale Netz trainiert und getestet wurde, wird ausgewertet, ob die ausgewählte Architektur für eine Situationserkennung auf humanoiden Robotern geeignet ist
Atomic Layer Deposition of the Conductive Delafossite PtCoO2
No full textThe first atomic layer deposition process for a ternary oxide is reported, which contains a metal of the platinum group, the delafossite PtCoO2. The deposition with the precursors trimethyl-Pt-methylcyclopentadienyl, Co-bis(N-t-butyl-N′-ethylpropanimidamidate), and oxygen plasma results in a process with a nearly constant growth rate and stoichiometric composition over a wide temperature window from 100 to 320 °C. Annealing of the as-deposited amorphous films in an oxygen atmosphere in a temperature window from 700 to 800 °C leads to the formation of the delafossite phase. Very thin films show a pronounced preferred orientation with the Pt sheets being almost parallel to the substrate surface while arbitrary orientation is observed for thicker films. The conformal coating of narrow trenches highlights the potential of this atomic-layer-deposition process. Moreover, heterostructures with magnetic films are fabricated to demonstrate the potential of PtCoO2 for spintronic applications